Three years after the launch of the Global Alliance for Genomics and Health (GA4GH) and six months after the first GA4GH-hosted convention of national genomics initiatives, Kathryn North (Australian Genomics) and Genomics England’s Chief Scientist, Professor Mark Caulfield recently convened representatives from 13 National Initiatives in genomic data collection to discuss areas of potential collaboration at the Wellcome Trust in London.
The goal of the meeting was to identify potential areas of collaboration, resource and expertise sharing, as well as common needs across National Initiatives that GA4GH can incorporate into its “toolbox” of data sharing standards and tools.
The meeting was attended by representatives of major population-sequencing endeavours, from: Australia, Brazil, Canada, Finland, GenomeAsia100k, Global Gene Corp (India), the Netherlands, Qatar, South Africa, Switzerland, Turkey, the USA and the UK.
The three more established projects – Genomics England, Australia and the USA gave detailed talks about progress to date and challenges they have faced. This was followed by presentations from all of the emerging projects – covering their aims, funding models, disease focus, and resources they’ve developed. There were also workshops on Data & Sharing, Regulation, and Clinical & Education, as well as a round-table discussion.
Sir John Chisholm, Executive Chair of the Genomics England Board, spoke on his vision for genomic research:
"This change is possible because we will soon be able to understand the genome, and to use it to predict outcomes.
“It’s a fantastic vision, but it’s very hard and it will take most of rest of the century to get there,” said Chisholm. He cited two challenges in particular:
1. Genomic medicine involves “colossal” amounts—millions, tens of millions, or even 100s of millions—of data points. This scale is necessary because of the low probability of making connections between the genome and human health.
Additionally, because very few associations are monogenic, combinatorial problems make unpacking the genome a very difficult and complex pursuit. In the past, scientific programs have built individual research cohorts for each study. This will not work with clinical genomic research, because no organization has enough money to fund research cohorts at this scale.
The only way to achieve cohorts of the size needed is by aligning fully consented patient data from the healthcare system with genomic data. No one country will be able to do this alone so national programs must collaborate.
Given that no one country can do it alone, nations must agree on rigorous standards and protocols, as artefacts of non-harmonized data collection processes will make it impossible to understand outcomes across a combined data set.
2. Implementing standards and protocols in routine healthcare will be difficult, Chisholm said, “but it’s something where the prize is so great, it’s worth doing.” In the four years since it was launched, Genomics England has spent considerable effort and made progress on the development of standards and protocols for data collection and getting them implemented across the UK’s National Health Service.
Now, he said, that needs to be taken to the international stage. He invited meeting attendees to “form a club” to work together to agree on standards and protocols for clinical genomic data sharing. This will allow for federated data sharing that is protective of participant confidentiality and privacy and enables “this transformation of the human experience in the 21st century.”
In closing remarks, Professor Mark Caulfieldsaid that the groups present must showcase their achievements to patients and the publicin order to promote further involvement from those communities. Doing so, he said will result in “more ‘Denmarks’ and less places where we lock data away.”
He also summarised a number of high level themes emerging from the meeting:
- In three years, the discussion has moved from a focus on future goals to examples of established infrastructure and large-scale genomic data collection in health care systems around the world
- The next step is to position healthcare against research in order to achieve the needed scale, and to move from cohorts of 1 million individuals to 5 or 10 million thanks to shared data across the globe
- This will require an ethical framework and robust guidelines for follow up
- Initiatives must engage with patient advocacy and support groups since those populations are willing to take on challenges that researchers shirk from
- Industry is vital to this endeavor, as it requires millions (or billions) of dollars
- There is a need for robust IP that allows for freedom to operate, and the ability to account for situations where the responsible act is not to protect IP but to give it away